• The Vendor's Bazaar forum is for commercial offers and transactions. Only unmoderated members can post here.

    diyAudio provides this forum for the convenience of our members, but makes no warranty nor assumes any responsibility. We do not vet any members. Use of this facility is at your own risk. Customers can post any issues in those threads as long as it is done in a civil manner. All diyAudio rules about conduct apply and will be enforced.

Output transformers for DACs

Jump to Latest
#831
Hi All,

I'm working on a streamer/DAC project using Ian Canada's parts and in particular the 9038Q2MPi Dual Mono Plus DAC HAT (389 Ohm output impedance, 1.4 mA rms ?). I'm looking at the output stage which led me to Ivan's transformers. After an initial email exchange with Ivan, I thought my following questions might be relevant to other people and worth putting in the chat, at the risk of exposing myself as a complete newbie. The following is based on my readings over the last few weeks, so please feel free to correct me if there are mistakes. The gist of it is I'm trying to understand the parameters of transformers, i.e. what parameters are relevant for my project and how they interact: turns ratio, primary and secondary resistors, adding capacitors?

Factors affecting my build and thinking process:
  • Passive output stage only: I'm trying to keep the first iteration of my project to a small form factor, hence only having 3.3V and 5V PSUs and no 15V as I'll be using Ian Canada PurePi 5V+3.3V. This would lead to having a passive transformer output stage only, or potentially an OPA output stage using low voltage OPA that could use the 5V PSU, like the OPA861
  • Swappable output stage:
    1. Part of the joy of this DIY project is tinkering and trying different options, so I'd like to have different output stages that I can mount and see how they affect the sound. As the output stage sits at the top should be easy to swap them
    2. Different usage cases. I'm planning to use my streamer/DAC in 2 different situations: i) going into a tube preamp with a 600 Ohm input impedance and ii) going directly into various amps with input impedance ranging from 30K Ohm to 100K Ohm. So again, I thought being able to swap the output stage might be useful to match these use cases better
From going through this thread and other forums, my very basic understanding is:
  • Output stage is there to match the DAC output impedance to the input impedance of the next item in the chain, amp or preamp
  • Ideally you need the DAC to "see" an impedance of 20 Ohm maximum, but ideally as low as possible. I guess this will be impacted by the turn ratio of the transformers as well as the primary and secondary resistors, but I cannot seem to find the maths to help me decide which values to pick for those to make it work?
  • Is it a mistake to try and keep the output stage passive as the only solution to make it work with the tube preamp with 600 Ohm input impedance is to make it active and use jfets?
Based on the above, I was considering having the following hats to swap around:
  1. Ian Canada I/V STD mkII OPA I/V stage with OPA861. Can swap OPAs later on if I upgrade the setup to have a 15V PSU. Will it work for both my tube preamp at 600 Ohm input and my tube amp at 100k Ohm input?
  2. Ivan's transformers, but only for high input impedance amps (tube amp at 100K Ohm or class D amp at 47K Ohm), as it would need to be active to work with 600 Ohm input impedance
  3. Ian Canada's transformer I/V kit with 600:600 transformers to use with my tube preamp at 600 Ohm input impedance
  4. Ian Canada's transformer I/V kit with 600:50k transformers to use with my tube amp at 100K Ohm input impedance. This is based on my very basic understanding that we are trying to match impedance, but no idea what resistors to use?
  5. LL1544A seem to be an interesting choice as well, as you seem to be able to use it in different scenarios as per the table below. But again, I don't understand the circuit logic and maths behind how changing the turns ratio and primary and secondary resistance will impact your impedance matching use case and THD


Any advice, contribution, or pointing to extra literature on the subject would be very helpful

Thanks!
 
#832
Hi grarno,

Hope my comments will help you somehow.
grarno said:
389 Ohm output impedance, 1.4 mA rms ?
Click to expand...
The output current of Ian's DAC is 2.7mA (1.369*2) as it is a dual mono configuration.
grarno said:
Output stage is there to match the DAC output impedance to the input impedance of the next item in the chain, amp or preamp
Click to expand...
The output stage is for current to voltage conversion as ES90x8 poorly works in high-impedance (voltage-type) mode in terms of SQ. Impedances matching with the next stage is needed for sure as well, but in case of OPA approach the output impedance is low enough for say 98% cases of the next stages. In case of SUTs (my transformers for ES90x8 are step-up) the output impedance is suitable for say 90% of cases.
grarno said:
Ideally you need the DAC to "see" an impedance of 20 Ohm maximum, but ideally as low as possible. I guess this will be impacted by the turn ratio of the transformers as well as the primary and secondary resistors, but I cannot seem to find the maths to help me decide which values to pick for those to make it work?
Click to expand...
Correct. There are two versions of my P'n'P boards available for Ian's DM9038q2mpi DAC. On smaller trafos with turns ratio (TR) of 1:8 and on a larger trafos with turns ratio of 1:16. Both trafos works and both sounds better than OPAs (it has checked for many times). Larger trafos are better. They was better even at the same TR, but after some probes I have stopped at 1:16 for larger trafos.
Regarding maths. For 1:16 trafos I use 75R at the primary side and 100k+100k at the secondary side.
DAC sees 75R || (100k+100k)/(16*16) =>
=> 75R||~800R=69R.
The output impedance after 1:16 trafos is:
69R*(16*16)=18k
Output level: 69*0.0027*16=~3Vrms

The most often case is the next stage (preamp or amp) with the input impedance of 50k. So the corrected values will be:
DAC sees: 75R||800R||195R=51R
Combined (output||input impedance): 51R*256=13K
Output level: 51*0.0027*16=~2.2Vrms

grarno said:
Is it a mistake to try and keep the output stage passive as the only solution to make it work with the tube preamp with 600 Ohm input impedance is to make it active and use jfets?
Click to expand...
It depends from the needed signal level (voltage) at the input of your 600R preamp. 600R will be reflected from the secondary coil and DAC will see

600/256=2.3R. And as a result you will get 0.1Vrms of the signal only. If it is OK input level for your preamp and further chain then cool. Situation depends on the amplification factor of your preamp I suppose. But sure, usage of the current buffer like an active stage on 2sk170/2sj74 jfet pairs possible on my PCB after trafos will give you more suitable voltage output at a low impedance almost in 100% cases of the next stage input requirements.

grarno said:
Ian Canada I/V STD mkII OPA I/V stage with OPA861. Can swap OPAs later on if I upgrade the setup to have a 15V PSU. Will it work for both my tube preamp at 600 Ohm input and my tube amp at 100k Ohm input?
Click to expand...
I think yes as the output impedance after OPA should be low enough
grarno said:
Ivan's transformers, but only for high input impedance amps (tube amp at 100K Ohm or class D amp at 47K Ohm), as it would need to be active to work with 600 Ohm input impedance
Click to expand...
grarno said:
Ian Canada's transformer I/V kit with 600:600 transformers to use with my tube preamp at 600 Ohm input impedance
Click to expand...
Good, but DAC will work at a voltage-type output mode (DAC will see 600R of load). Difference will be in SQ of trafos itself.

grarno said:
Ian Canada's transformer I/V kit with 600:50k transformers to use with my tube amp at 100K Ohm input impedance. This is based on my very basic understanding that we are trying to match impedance, but no idea what resistors to use?
Click to expand...
Similar to my approach. Difference will be in SQ of trafos itself.
grarno said:
LL1544A seem to be an interesting choice as well, as you seem to be able to use it in different scenarios as per the table below. But again, I don't understand the circuit logic and maths behind how changing the turns ratio and primary and secondary resistance will impact your impedance matching use case and THD
Click to expand...
Guys has compared with my trafos many times. Apart from different possible configuration of TRs, as per customers feedbacks my trafos (even smaller type) are better. Finaly the difference will be in SQ of trafos itself.

Thank you.
Ivan.
 
  • Like
Reactions: miklau
#839
Here's the DAC I built about ten years ago with the Lundahl xformers. It was designed by a friend of mine. I have a second one that has a discreet opamp output stage in place of the Xformers. The Xformer version however is the better sounding one of the two.


DAC Ver 1.0.jpg
 
#840
Mark Allen said:
Here's the DAC I built about ten years ago with the Lundahl xformers. It was designed by a friend of mine. I have a second one that has a discreet opamp output stage in place of the Xformers. The Xformer version however is the better sounding one of the two.


View attachment 1115904
Click to expand...
Why are you posting this here - this is a vendor thread in the vendor bazaar. If you want to show people your DAC you should start your own thread or perhaps post in the DAC Gallery thread.
 
  • Like
Reactions: MikeyFresh
You must log in or register to reply here.
Top Bottom